| Abstract: | Previous studies have demonstrated the preventive effects of greenness and greenspace using three classes of exposure metrics (greenspace coverage, greenness, and landscape indices for greenspace) on suicide. Nevertheless, most studies were individually employed for each metric using a single-phase method, representing the short-term exposure. We hierarchically investigated the association between three classes of exposure metrics and suicide using long-term data from 2000 to 2018 in Taiwan. Normalized difference vegetation index (NDVI) was used to represent the greenness exposure, while mean patch area (area and edge), area-weighted mean fractal dimension index (shape), and proximity index (proximity) were opted to represent landscape indices for greenspace. Using the generalized additive mixed model, we revealed that the relative risk (95% confidence interval) of greenspace coverage, NDVI, mean patch area, area-weighted mean fractal dimension index, and proximity index was 0.998 (0.997, 0.999), 0.860 (0.811, 0.912), 0.999 (0.999, 0.999), 1.485 (1.056, 2.088), and 0.999 (0.999, 0.999), respectively. This indicates that long-term exposure to both greenspace coverage and greenness is associated with a reduced suicide ratio. Similarly, long-term exposure to greenspace that is greater in size, simpler in shape, and closer in proximity is associated with a decreased suicide risk. For formulating future afforestation policies, the layout and connectivity of greenspaces should be considered besides increasing green areas. This study not only demonstrated the significance of long-term exposure to greenspace for reducing the suicide ratio, but also illustrated a hierarchical approach to examine the effects of greenspace. |
